INDEX

Explanations

mentions of the letter "A" with high activation values

oai_token-act-pair · gpt-3.5-turbo

the letter 'A' in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Vaugh

-0.67

enegger

-0.65

 Alph

-0.63

 Indigo

-0.60

reports

-0.60

email

-0.59

 Nicaragua

-0.58

 optics

-0.57

 Everton

-0.57

Param

-0.57

POSITIVE LOGITS

cknowled

1.19

verages

1.18

ussie

1.17

cknow

1.11

uctions

1.10

irst

1.01

lyss

0.98

roma

0.98

ryan

0.96

perture

0.94

Activations Density 0.105%

mentions of the letter "A" with high activation values

the letter 'A' in various contexts

No Comments

No Known Activations